Method and System for Adaptive Direct Volume Rendering

1. A method of adaptive direct volume rendering, comprising: fragmenting a sampled 3-D dataset of a scalar field into a plurality of sub-volumes of different sizes, each sub-volume associated with a set of data value parameters characterizing the data value distribution of the scalar field within the sub-volume; defining an opacity transfer function that is dependent upon data values of the scalar field and an illumination model; selectively casting a plurality of rays from a 2-D image plane towards the sampled dataset using a computer processor, each ray having an initial ray energy and a cross-section; for each ray cast from a selected location on the 2-D image plane, selecting a subset of the plurality of sub-volumes for interaction with the ray; estimating the ray energy reflected by each sub-volume of the subset using the opacity transfer function and the illumination model; and summing the reflected ray energy as a pixel value at the selected location on the 2-D image plane; and estimating pixel values at other locations on the 2-D image plane using the pixel values at the selected locations wherein the step of estimating the ray energy reflected by each sub-volume of the subset includes; estimating a maximum energy differential of the sub-volume; comparing the maximum energy differential against a predefined energy error threshold; if the maximum energy differential is above the predefined energy error threshold, recursively selecting a smaller sub-volume along the ray path; and estimating a new maximum energy differential of the smaller sub-volume ; and if the maximum energy differential is below the predefined energy error threshold, calculating the amount of ray energy reflected by the sub-volume using the illumination model wherein if the sub-volume is a smallest sub-volume comprising 2×2×2 3-D cells, the smaller sub-volume is a 3-D cell within the smallest sub-volume, and if the opacity transfer function varies monotonically within the cell and an iso-surface exist in the 3-D cell, the maximum energy differential of the 3-D cell is calculated using the eight data values at the corners of the 3-D cell and the opacity transfer function.

2. The method of claim 1 , wherein the maximum energy differential depends on the opacity transfer function and the maximum, average, and minimum data values of the sub-volume.

3. The method of claim 1 , wherein the amount of ray energy reflected by the sub-volume depends on the length of ray path within the sub-volume, the opacity transfer function within the sub-volume the average scalar field value of the sub-volume, and the local gradient vector of scalar field within the sub-volume.

4. The method of claim 1 , wherein if the sub-volume is a smallest sub-volume comprising 2×2×2 3-D cells, the smaller sub-volume is a 3-D cell within the smallest sub-volume, and if the opacity Function does not vary monotonically within the cell, the 3-D cell is further divided into multiple sub-cells until the dimension of a smallest sub-cell reaches the cross-section of the ray.

5. The method of claim 1 , wherein if the sub-volume is a smallest sub-volume comprising 2×2×2 3-D cells, the smaller sub-volume is a 3-D cell within the smallest sub-volume, and if the opacity function varies monotonically within the cell, the maximum energy differential of the 3-D cell is estimated by dividing the maximum energy differential of the sub-volume by 2.

6. The method of claim 1 , wherein the predefined energy error threshold is modulated by an image rendering speed specified by a user and a zoom factor in the case of parallel projection or a perspective angle and a perspective distance between the image plane and the 3-D dataset in the case of perspective projection.

7. A method of adaptive direct volume rendering, comprising: fragmenting a sampled 3-D dataset of a scalar field into a plurality of sub-volumes of different sizes, each sub-volume associated with a set of data value parameters characterizing the data value distribution of the scalar field within the sub-volume; defining an opacity transfer function that is dependent upon data values of the scalar field and an illumination model; selectively casting a plurality of rays from a 2-D image plane towards the sampled dataset using a computer processor, each ray having an initial ray energy and a cross-section; and for each ray cast from a selected location on the 2-D image plane, selecting a subset of the plurality of sub-volumes for interaction with the ray wherein the subset constitutes a largest sub-volume in which the optical transfer function is determined to vary monotonically between maximum and minimum data values; estimating the ray energy reflected by each sub-volume of the subset using the opacity transfer function and the illumination model; and summing the reflected ray energy as a pixel value at the selected location on the 2-D image plane.

8. The method of claim 7 wherein the step of selecting a subset of the plurality of sub-volumes for interacting with the ray includes: identifying a largest sub-volume along the ray path and its corresponding maximum and minimum data values; checking if the opacity transfer function varies monotonically between the maximum and minimum scalar field values; if the function does not vary monotonically, recursively identifying a smaller sub-volume along the ray path and its corresponding maximum and minimum data values; and checking if the opacity transfer function varies monotonically between the maximum and minimum scalar field values of the smaller sub-volume; and if the function does vary monotonically, estimating the amount of ray energy reflected by the sub-volume during its interaction with the ray.

9. The method of claim 7 , wherein two lookup tables are constructed for the opacity transfer function such that a forward lookup table contains the data value difference to a nearest local extreme of the opacity transfer function along the data value increasing direction and a backward lookup table contains the data value difference to a nearest local extreme of the opacity transfer function along the data value decreasing direction.

10. The method of claim 9 , wherein if the maximum data value of the sub-volume is smaller than the summation of the minimum data value of the sub-volume and its corresponding data value difference stored in the forward lookup table or the minimum data value of the sub-volume is larger than the difference between the maximum data value of the sub-volume and its corresponding data value difference stored in the backward lookup table, the opacity transfer function varies monotonically between the minimum and maximum data values.

11. A method of adaptive direct volume rendering, comprising: fragmenting a sampled 3-D dataset of a scalar field into a plurality of sub-volumes of different sizes, each sub-volume associated with a set of data value parameters characterizing the data value distribution of the scalar field within the sub-volume; defining an opacity transfer function that is dependent upon data values of the scalar field and an illumination model; and selectively casting a plurality of rays from a 2-D image plane towards the sampled dataset using a computer processor, each ray having an initial ray energy and a cross-section; and for each ray cast from a selected location on the 2-D image plane, identifying a largest sub-volume along the ray path in which the optical transfer function varies monotonically between maximum and minimum data values; estimating the ray energy reflected by the largest sub-volume using the opacity transfer function and the illumination model; and summing the reflected ray energy as a pixel value at the selected location on the 2-D image plane.

12. The method of claim 11 wherein the step of identifying the largest sub-volume along the ray path in which the optical transfer function varies monotonically includes: identifying a largest sub-volume along the ray path and its corresponding maximum and minimum data values; checking if the opacity transfer function varies monotonically between the maximum and minimum scalar field values; if the function does not vary monotonically, recursively identifying a smaller sub-volume along the ray path and its corresponding maximum and minimum data values; and checking if the opacity transfer function varies monotonically between the maximum and minimum scalar field values of the smaller sub-volume; and if the function does vary monotonically, estimating the amount of ray energy reflected by the sub-volume during its interaction with the ray.

13. The method of claim 11 , wherein two lookup tables are constructed for the opacity transfer function such that a forward lookup table contains the data value difference to a nearest local extreme of the opacity transfer function along the data value increasing direction and a backward lookup table contains the data value difference to a nearest local extreme of the opacity transfer function along the data value decreasing direction.

14. The method of claim 13 , wherein if the maximum data value of the sub-volume is smaller than the summation of the minimum data value of the sub-volume and its corresponding data value difference stored in the forward lookup table or the minimum data value of the sub-volume is larger than the difference between the maximum data value of the sub-volume and its corresponding data value difference stored in the backward lookup table, the opacity transfer function varies monotonically between the minimum and maximum data values.